genesis of carbonate aggregates in lamprophyres from the ...azbej_min_pet.pdf · virginia tech....

Mineralogy and Petrology (2006) 88: 479–497DOI 10.1007/s00710-006-0123-y

Genesis of carbonate aggregatesin lamprophyres from the northeasternTransdanubian Central Range, Hungary:Magmatic or hydrothermal origin?

T. Azbej1;2, C. Szabo1, R. J. Bodnar2, and G. Dobosi3

1 Lithosphere Fluid Research Laboratory, Department of Petrologyand Geochemistry, E€ootv€oos University, Budapest, Hungary2 Fluids Research Laboratory, Department of Geosciences,Virginia Polytechnic Institute and State University, Blacksburg, USA3 Institute for Geochemical Research, Hungarian Academy of Sciences,Budapest, Hungary

Received May 18, 2005; revised version accepted January 18, 2006Published online May 11, 2006; # Springer-Verlag 2006Editorial handling: B. DeVivo

Summary

Carbonate aggregates in Late Cretaceous lamprophyre dikes of the northeasternTransdanubian Central Range (TCR) in Northwest Hungary have been classified intothree genetic groups. Type-I dolomiteþ calcite� magnesite aggregates have petrograph-ic and geochemical features similar to ocelli described by other workers. Fluid inclusionsin Type-I aggregates homogenize between 77 and 204 �C and are of hydrothermal origin.Type-II aggregates are characterized by a polygonal shape and are mostly dolomite.Based on their shape and primary fluid inclusions which homogenize between 95 and172 �C, these carbonate aggregates are interpreted to fill vugs produced by the dissolu-tion of olivine phenocrysts. Type-III carbonate aggregates show an irregular to polygonalshape and distinct compositional zonation and contain secondary aqueous fluid inclu-sions. Homogenization temperatures of fluid inclusions are below 104 �C, and zonationpatterns suggest partial recrystallization. These carbonate aggregates are most likelyxenoliths and xenocrysts from the wall rocks of the lamprophyre melt conduits.

Introduction

An igneous rock is said to have an ocellar texture if the ‘‘phenocrysts’’ (ocelli)consist of aggregates of smaller crystals arranged radially or tangentially around

larger, generally euhedral crystals to produce a rounded eyelike appearance (e.g.Phillpotts, 1990). Ocelli of globular, carbonate or felsic silicate aggregates are com-mon features of lamprophyres and some alkali basalts (e.g. Phillpotts, 1990; Rock,1991). For decades ocelli in igneous rocks were interpreted to be the products ofsilicate–carbonate or silicate–silicate liquid immiscibility or, rarely, as amygdales orvesicles filled by late stage minerals (e.g. Phillpotts and Hodgson, 1968; Fergusonand Currie, 1971; Hamilton et al., 1979; Cooper, 1979; Foley, 1984; N�eedli and M. Toth,2003; Vichi et al., 2005). Ocelli are not the only occurrence of carbonate minerals inlamprophyres: carbonates are also found as pseudomorphs after olivine, melilite orother minerals, as intergrowths with talc, garnet, etc., and as late veins (Rock, 1991).Rock (1991) classifies primary and secondary carbonates based on stable isotopiccomposition of lamprophyres from locations around the world.

Carbonate and silicate ocelli have been described in Late Cretaceous lampro-phyre dikes in the Transdanubian Central Range (TCR) in Northwest Hungary (e.g.Horvath et al., 1983; Szabo et al., 1993). The ocelli have been interpreted to be theresult of liquid immiscibility in volatile-rich mafic melts (Kubovics et al., 1990).Based on the stable isotopic composition, Dem�eeny et al. (1994) showed that theocelli from TCR lamprophyre dikes represent a transition between primary igneousand sedimentary carbonate. They interpreted the isotopic composition of the aggre-gates to be the result of hydrothermal fluids, suggesting a genetic model incon-sistent with melt immiscibility. Later detailed petrographic, geochemical and fluidinclusion studies of carbonate aggregates (including ocelli) from the TCR lampro-phyres (Azbej, 2002; Azbej et al., 2003) called into question an origin by meltimmiscibility or by simple mixing of different fluids. Because the term ‘‘ocelli’’has genetic implications, we will use the more general term ‘carbonate aggregates’to describe these features.

The goal of this study is to constrain the genesis of the various types of carbon-ate aggregates observed in the TCR lamprophyres. Petrographic, cathodolumines-cence and electron microprobe analyses of the carbonate aggregates and fluidinclusion microthermometry provide information related to the origin of carbonateaggregates, as well as the late-stage evolution of the TCR lamprophyres.

Geological background and lamprophyre petrology

During the 1980s, unusual alkaline mafic lamprophyre dikes were recognized inquarries in the NE Transdanubian Central Range, Hungary (Horvath and �OOdor,1984; Kubovics et al., 1989), and in several boreholes (Horvath et al., 1983; Szabo,1985; Kubovics and Szabo, 1988; Kubovics et al., 1989) (Fig. 1). These rocks areLate Cretaceous in age (Horvath et al., 1983; Embey-Isztin et al., 1989; Dunkl,1991) and occur as dikes or dike swarms and show consistent petrographic andgeochemical features including panidiomorphic granular texture, high color index,ocelli content, high volatile and incompatible element abundances and similar ages(Szabo et al., 1993). Clinopyroxene and olivine occur as phenocrysts in the lam-prophyres, and many of the olivine grains are altered. The groundmass is charac-terized by a framework of clinopyroxene and mica microphenocrysts amongcarbonate, feldspar, apatite and glass. Multiple generations of carbonate-filledveins are also present. Based on modal composition, most of the dike-rocks are

480 T. Azbej et al.

Fig. 1. Geological map of the northeastern Transdanubian Central Range (TCR). Blackcircles show locations of lamprophyres in outcrop and drill holes (from Szabo et al., 1993).Fluid inclusion and chemical data on host phases were obtained from outcrop samples Paand Rh, and drill core samples Ad-2, Bkt-1, B€oo-1, My-1, St-1, Val-3; inset map shows theCarpathian-Pannonian Region and the study area [black box] (Azbej, 2002)

Genesis of carbonate aggregates in lamprophyres 481

classified as monchiquite, but camptonite and alnoite also occur (Embey-Isztinet al., 1989; Kubovics and Szabo, 1988; Kubovics et al., 1989; Szabo et al.,1993). Some dikes contain many upper mantle to crustal xenoliths and xenocrysts(Szabo, 1985; Kubovics et al., 1985, 1989). The dikes are emplaced into Carbonif-erous granites, Upper Permian sandstones, Triassic carbonates and andesites.Some carbonatite-like dikes classified as beforsite are found in the St-1 borehole(Horvath et al., 1983). The St-1 samples are composed of dolomite (occurring ascarbonate aggregates), phlogopite and plagioclase as dominant phases (Horvathet al., 1983). Another carbonatite-like dike containing mostly calcite and phlogo-pite has been described in the Ad-2 borehole (Szabo et al., 1993).

Geochemical characteristics suggest that the lamprophyric melt was intrudedinto an intra-plate extensional tectonic zone (Kazm�eer and Szabo, 1989). Accord-ingly, lamprophyres are typical of Alpine nappe units located at the outer edge ofthe accretionary wedge. During the Cretaceous these Alpine nappe units were partof a continental basement far away from the inferred Penninic subduction zone(Kazm�eer and Szabo, 1989; Kubovics et al., 1989). The most likely origin of theparental lamprophyric melt is through a small degree of partial melting of meta-somatized lithospheric mantle (Szabo, 1985; Szabo et al., 1993).

Samples studied and analytical methods

Samples were collected from 11 dikes exposed in 2 surface outcrops (4 represen-tative samples: Pa=1, Pa=2, Rh=121, Rh=1012) and 6 boreholes (nine representa-tive samples: Ad-2=II, Ad-2=III, Ad-2=VIII, Ad-2=X, Bkt-1=6, B€oo-1=16, My-1=1,St-1=1 Val-3=2) (Fig. 1). In addition, data from 6 samples collected by Dem�eenyet al. (1994) (Ad-2, Bkt-1, B€oo-1, Pa, Rh, St-1) have been incorporated into thispresent study.

Electron microprobe analyses were carried out on a JEOL Superprobe JXA-8600 WDS at the Department of Earth Sciences, University of Florence and on aCameca SX-50 at the Department of Geosciences at Virginia Tech. The accelerat-ing voltage was 15 kV, with a 10 nA sample current. Mineral analyses were per-formed using a beam diameter of 10 mm. Counting times were 20 seconds for allelements. Natural mineral standards were employed, and the correction method ofBence and Albee (1968) was applied at the University of Florence, and the ZAFtechnique was used at Virginia Tech. Backscattered electron images were collectedusing a JEOL Superprobe 733, equipped with an INCA Energy 200 EDS, at theInstitute for Geochemical Research of the Hungarian Academy of Sciences usingan accelerating voltage of 20 kV with a 3 nA sample current.

Microthermometric analyses were carried out on a Linkam THMS 600 stage atVirginia Tech. Cathodoluminescence studies were carried on a Technosyn 8200MK II cold cathodoluminescence microscope at Virginia Tech using a samplecurrent of 600 nA and an accelerating voltage of 10 kV.

Petrography

Szabo et al. (1993) reported ocelli from the TCR lamprophyres that containedcalcite and dolomite, radial and acicular micas, sanidine, analcime, and oxide

482 T. Azbej et al.

minerals, but these authors did not subdivide the ocelli into different types based ontextures. In the present study, three textural groups of carbonate aggregates weredistinguished. The volume of carbonate aggregates in the lamprophyres varies fromabout 1 to 45%.

The temporal classification of fluid inclusions in the carbonate aggregates ofthe TCR lamprophyres was determined based on petrographic characteristics of theinclusions (Goldstein and Reynolds, 1994; Bodnar, 2003a). Inclusions that occuralong fractures cross-cutting mineral grains were classified as secondary. Inclu-sions that appeared to be isolated and not along obvious fractures were interpretedto be primary.

Type-I aggregates have petrographic features characteristic of typical ocelli(Fig. 2A) and show a globular shape and are approximately 0.2 to 4.0 mm indiameter. In addition to carbonate minerals, silicates (nepheline, analcime, sani-dine) and rarely chalcedony are present either as anhedral phases occurring in theinterior, or as elongated subhedral phases intersecting the rim of the aggregates(Fig. 2B). In some samples phlogopite flakes occur tangentially at the rim of thecarbonate aggregates (Fig. 2C). The carbonate minerals contain primary fluidinclusions.

Type-II aggregates are found mostly in samples from the St-1 and rarely Ad-2dikes. The size varies from 0.2 to 3.5 mm in diameter. Their shapes and sizes aresimilar to those of olivine phenocrysts in the TCR lamprophyres (Fig. 2D). Type-IIaggregates contain minor chalcedony and primary fluid inclusions. Some thin sec-tions reveal multiple generations of carbonate veins intersecting each other andType-I and Type-II carbonate aggregates, forming a network (Fig. 2G, H).

Type-III aggregates are irregular to polygonal in shape (Fig. 2E, F), and varybetween 0.1 and 3.0 mm in diameter. In addition to the dominant carbonatephases, clay minerals sometimes occur near the rim (Fig. 2F). Cathodolumines-cence and SEM images reveal chemical zonation in these aggregates (Fig. 2F) incontrast to the relatively homogenous patterns observed in Type-I and Type-IIaggregates (Azbej, 2002). Type-III carbonate aggregates contain secondary fluidinclusions.

Chemical composition of carbonate aggregates

Several carbonate aggregates in each of seventeen samples from eight samplelocalities (Fig. 1, Fig. 3A, B; Tables 1, 2) have been analyzed. Type-I aggregatesconsist of calcite, dolomite and minor magnesite (Table 1) and show no com-positional or petrographic zoning or evidence of exsolution. Type-II aggregatesare composed of ferroan dolomite with FeO of 0.5 to 9.1 wt% and SrO up to1.7 wt% (Table 2).

A diagnostic characteristic of Type-III aggregates is their zoned texture(Fig. 2E, F). Based on electron microprobe traverses, a trend from calcite!dolomite ! magnesite or from dolomite ! magnesite is observed from the core tothe rim in several Type-III aggregates. The thickness of individual zones is on theorder of 100 mm. Sometimes ankerite is also observed at the rims. Similar compo-sitional zonation has been reported from the Rh sampling locality by Dem�eeny et al.(1994) where a carbonate aggregate with calcite core and dolomite rim was found.


484 T. Azbej et al.

Dem�eeny et al. (1994) also mention zoned aggregates containing dolomite andmagnesite from lamprophyres in the Ad-2 borehole.

All aggregates have relatively high trace and minor element concentrationcompared to the composition of the Triassic carbonate wall-rocks presented byBalog et al. (1999). Carbonate veins have been analyzed and compared to thecomposition of carbonates in coexisting aggregates (Table 3). In each of the sam-ples the composition of veins showed similar variation as of the aggregate carbon-ates. The composition of vein filling carbonates varies from predominantly calcitic(e.g. Ad-2) to ankeritic (B€oo-1=16, St-1) (Table 3).

Microthermometric analyses

The formation temperature of carbonate aggregates in the TCR lamprophyreswas estimated based on microthermometric analyses of fluid inclusions andchemical compositions of coexisting silicate phases in predominantly carbonateaggregates.

1Fig. 2. Photomicrographs (A, C, D, E, G, H) and BSE images (B, F) of carbonate aggregatesin the TCR lamprophyres. A Type-I aggregate containing calcite (CC) in sample St-1=1(plane-polarized light). B Type-I silicate-bearing aggregate containing calcite (CC) andnepheline (NE) in sample My-1=1. C Type-I aggregate containing dolomite (DO) sur-rounded by tangentially aligned phlogopites (PH) in sample St-1=1 (plane-polarized light).D Type-II aggregate containing dolomite (DO) in sample St-1=1 (plane-polarized light).E Type-III aggregate containing calcite (CC) and dolomite (DO) with glassy rim (GR) insample B€oo-1=16 (plane-polarized light). F Mineralogically zoned Type-III aggregatecontaining calcite (CC), dolomite (DO), magnesite (MG) and Al-silicate (SIL) in sampleRh=1012. G Two generations of cross-cutting carbonate veins (VN1, VN2) in sample Ad2-II. H Carbonate vein (VN) cross cutting a Type-II dolomite aggregate (DO) in sample St-1

Fig. 3. Ternary CaO–MgO–FeO diagrams showing compositions of carbonate phases inaggregates from the TCR lamprophyres. A Type-I, Type-II and Type-III aggregates insamples from outcrops (Pa, Rh) and drill cores (Ad-2, Bkt-1, B€oo-1, My-1, St-1, Val-3).B Composition of carbonates in samples Ad-2=6, Ad-2=7 and Rh=8 after Dem�eeny et al.(1994)


Tab

le1

.A

vera

ge

com

posi

tion

of

carb

onate

min

erals

inTyp

e-I

aggre

gate

sfr

om

the

TC

Rla

mpro

phyr

es

Ag

gre

gat

ety

pe

Ty

pe-

I

Sam

ple

Ad

-2=

III=

3B

kt1=

6M

y-1=1

My

-1=

4P

a1=

1P

a1=2

St-

1=

1S

t-1=2

Val

3=

2

Car

bo

nat

ep

has

eD

olo

mit

eM

agn

esit

eC

alci

teC

alci

teD

olo

mit

eM

g-C

alci

teC

alci

teC

alci

teC

alci

teF

e-D

olo

mit

eD

olo

mit

eD

olo

mit

eD

olo

mit

e

CaO

29

.40

.11

54

.85

5.4

34

.44

0.0

55

.05

4.2

55

.03

1.9

28

.72

9.6

28

.8

Mg

O2

1.2

46

.00

.12

0.0

51

3.5

9.9

90

.56

0.1

40

.21

15

.71

8.8

18

.92

0.3

FeO

0.4

93

.13

0.1

50

.12

3.9

23

.34

b.d

.0

.10

0.1

96

.01

4.0

44

.06

1.1

4

Mn

O0

.13

0.1

6b.d

.0

.02

0.1

50

.35

b.d

.0

.16

0.1

10

.61

0.3

10

.19

0.6

1

SrO

0.1

1�

b.d

.0

.36

0.3

50

.47

0.7

30

.45

0.2

40

.75

0.2

01

.01

0.2

42

.05

BaO

0.1

0�

0.1

2�

0.0

9�

0.1

0b.d

.b.d

.b.d

.0

.13�

0.0

9�

0.1

00

.09�

b.d

.0

.10�

To

tal

51

.24

9.4

55

.45

6.0

52

.55

4.4

56

.05

4.8

56

.25

4.5

52

.85

3.0

52

.9

Cal

c.C

O2

46

.64

9.6

43

.34

3.7

44

.54

4.9

44

.04

2.9

43

.94

6.3

46

.14

6.6

46

.7

Cal

c.T

ota

l9

8.0

99

.09

8.7

99

.79

7.0

99

.31

00

.09

7.8

10

0.1

10

0.8

98

.99

9.5

99

.6

CaC

O3

mo

l%5

4.1

0.3

19

9.0

99

.06

3.7

72

.19

8.1

98

.89

7.9

56

.75

2.1

53

.45

2.0

Mg

CO

3m

ol%

44

.78

9.3

0.1

30

.12

28

.72

0.7

1.1

30

.29

0.4

33

2.1

39

.23

9.2

42

.0

FeC

O3

mo

l%0

.83

10

.10

.07

0.2

16

.65

5.5

20

.03

0.1

70

.31

9.8

16

.69

6.7

31

.89

Mn

CO

3m

ol%

0.2

10

.25

–0

.03

0.2

60

.58

–0

.26

0.1

91

.00

0.5

20

.32

1.0

2

#o

fm

easu

rem

ents

72

34

11

25

34

73

3

� Hig

hes

tco

nce

ntr

atio

nm

easu

red

(not

aver

age)

for

ag

iven

trac

eel

emen

tin

the

sam

ple

b.d

.B

elow

det

ecti

on

lim

itC

alc

.C

O2

Cal

cula

ted

CO

2w

t%C

alc

.To

tal

Tota

lw

t%oxid

es(i

ncl

udin

gC

O2)

486 T. Azbej et al.

Tab

le2

.A

vera

ge

com

po

siti

on

of

carb

on

ate

min

era

lsin

Typ

e-II

an

d-I

IIa

gg

reg

ate

sfr

om

the

TC

Rla

mp

rop

hyr

es

Ag

gre

gat

ety

pe

Ty

pe-

IIT

yp

e-II

I

Sam

ple

Ad

-2=

XS

t-1=

3S

t-1=4

Ad

-2=II

bB€ oo1

-16=

1B€ oo1

-16=

2B€ oo1

-16=

3R

h1

01

2=

1R

h1

01

2=3

Car

bo

nat

ep

has

eD

olo

mit

eD

olo

mit

eF

e-D

olo

mit

eD

olo

mit

eC

alci

teD

olo

mit

eM

g-C

alci

teF

e-D

olo

mit

eA

nk

erit

eC

alci

teF

e-D

olo

mit

eM

agn

esit

e

Ag

gre

gat

ezo

ne

No

tzo

ned

No

tzo

ned

No

tzo

ned

core

core

rim

core

rim

rim

core

rim

rim

CaO

28

.82

9.4

29

.63

0.7

54

.71

1.2

43

.92

8.9

1.3

45

3.2

31

.81

.22

Mg

O1

8.9

17

.21

7.4

18

.60

.19

12

.04

.19

16

.53

2.8

1.1

51

8.0

40

.7

FeO

5.0

45

.78

6.3

43

.99

0.1

13

3.9

6.6

56

.69

18

.00

.98

2.5

47

.10

Mn

O0

.17

0.1

60

.20

0.0

90

.09�

0.6

30

.38

0.1

6b.d

.0

.53

0.4

70

.18

SrO

0.2

90

.08�

0.1

0�

0.1

50

.95

0.1

40

.35

0.1

0b.d

.b.d

.0

.12�

0.0

9�

BaO

b.d

.0

.09�

b.d

.0

.08�

b.d

.b.d

.0

.10

b.d

.b.d

.b.d

.0

.09�

b.d

.

To

tal

53

.35

2.6

53

.65

3.5

55

.95

7.9

55

.65

2.3

52

.15

5.8

52

.74

9.2

Cal

c.C

O2

46

.64

5.6

46

.34

7.0

43

.64

3.1

43

.44

5.0

47

.94

3.9

46

.44

9.9

Cal

c.T

ota

l9

9.9

98

.39

9.9

10

0.5

99

.61

01

.09

9.0

97

.31

00

.09

9.8

99

.39

9.2

CaC

O3

mo

l%5

1.7

53

.75

3.3

55

.09

8.0

19

.88

1.0

53

.42

.62

95

.15

7.5

2.2

4

Mg

CO

3m

ol%

39

.03

6.2

35

.93

8.1

0.4

02

4.3

6.9

03

4.8

68

.82

.36

37

.38

5.5

FeC

O3

mo

l%8

.66

9.6

71

0.4

6.5

10

.18

54

.71

1.5

11

.42

8.6

1.6

14

.21

11

.9

Mn

CO

3m

ol%

0.2

70

.27

0.3

30

.14

0.1

40

.97

0.6

30

.25

–0

.87

0.7

80

.29

#o

fm

easu

rem

ents

29

54

43

26

13

34

� Hig

hes

tco

nce

ntr

atio

nm

easu

red

(not

aver

age)

for

ag

iven

trac

eel

emen

tin

the

sam

ple

b.d

.B

elow

det

ecti

on

lim

itC

alc

.C

O2

Cal

cula

ted

CO

2w

t%C

alc

.To

tal

To

tal

wt%

ox

ides

(in

clu

din

gC

O2)


According to Boettcher et al. (1980), the water-saturated solidus of magmaticcarbonate is above 565 �C at pressures less than 5 kbars. Thus, carbonates of hydro-thermal origin would be expected to precipitate below 565 �C.

Primary fluid inclusions in Types-I and -II aggregates and secondary inclusionsin Type-III carbonates were analyzed. The carbonate veins did not contain fluidinclusions suitable for microthermometric analysis. Fluid inclusions in Types-Iand -II aggregates are scattered randomly in the host mineral, and show no align-ment along fracture planes or growth zones. These inclusions are interpreted to beof primary origin based on mode of occurrence. The inclusions vary from <2mmto about 14 mm, contain vapor and liquid at room temperature and have irregular toregular negative-crystal shape.

Type-III aggregates contain secondary fluid inclusions along roughly planar,sometimes branching, healed fractures that do not extend to the rim of the aggre-gates. The inclusions are <4 mm, contain liquid and vapor at room temperature,and have an irregular shape. The Type-III aggregates also contain a few randomlydistributed inclusions which would be classified as primary based on their occur-rence. However, the homogenization temperature and salinity of these inclusions issimilar to that of the obviously secondary inclusions. We interpret these randomlydistributed inclusions to be either secondary inclusions or primary inclusions thathave reequilibrated and=or refilled. This interpretation is based on the assumptionthat any primary inclusions in the carbonate that were trapped when it was origi-nally precipitated at depth would have reequilibrated when they were entrained intothe hot lamprophyric magma during transport to the near-surface (Bodnar, 2003b).

Table 3. Average composition of carbonate minerals in veins from the TCR lamprophyres

Sample Ad-2=II B€oo-1-16 St-1=5Carbonate phase Calcite Ankerite Fe-Dolomite

CaO 54.2 3.07 28.5MgO 0.71 32.3 9.83FeO 1.51 16.4 16.7MnO 0.18 0.24 0.11SrO 0.12� 0.13� 0.12BaO 0.10� 0.10� b.d.

Total 56.5 52.0 55.2

Calc. CO2 42.8 47.7 43.3Calc. Total 99.3 99.8 98.5

CaCO3 mol% 95.9 5.04 51.5MgCO3 mol% 1.74 73.6 24.7FeCO3 mol% 2.09 21.0 23.6MnCO3 mol% 0.25 0.31 0.15

# of measurements 3 7 2

�Highest concentration measured (not average) for a given trace element in the sampleb.d. Below detection limitCalc. CO2 Calculated CO2 wt%Calc. Total Total wt% oxides (including CO2)

488 T. Azbej et al.

Homogenization temperatures range from 77 to 204 �C (Fig. 4a, c) and areconsistent with data reported by Dem�eeny et al. (1994) on primary fluid inclu-sions in carbonate ocelli from some TCR lamprophyre dikes. However, our datashow a wider temperature range (Fig. 4a). Fluid inclusions in Types-I and -IIcarbonates span the entire temperature range mentioned above (Fig. 4a, b)

Fig. 4. Homogenization tempera-tures of fluid inclusions hostedin carbonate aggregates in TCRlamprophyres sorted accordingto the aggregate type. Data forType-I and Type-III aggregates in-clude results from Dem�eeny et al.(1994)


whereas fluid inclusions in Type-III carbonates vary within a narrower range(Fig. 4c).

First melting occurs at about �21 �C, near the H2O–NaCl eutectic temperature(�21.2 �C; Bodnar, 1992). No evidence of gases was observed. Final ice meltingtemperatures in the three aggregate types overlap and vary between �0.2 �C and�15.0 �C, with most between �2.4 �C and �9.6 �C (Table 4). Based on the ob-served first melting temperatures the inclusions are modeled using the H2O–NaClsystem. Salinities of primary fluid inclusions in Types-I and -II aggregatesrange from 4.0 to 13.6 NaClequiv. wt% (Bodnar, 1992), ignoring the extreme data(Table 2). Final melting temperatures of secondary inclusions in Type-III aggre-gates average �1.8 �C, corresponding to an average salinity of 2.8 NaClequiv. wt%.

Discussion

Geothermometry

The formation temperature of carbonates was estimated using the method ofHamilton (1961) for nepheline in equilibrium with carbonates. This method isbased on experimentally determined limits of NaAlSiO4–KalSiO4–SiO2–H2Osolid-solution in nepheline at temperatures between 500 and 775 �C, using theK=Na and Si=Al cation ratios of nepheline in equilibrium with silicate melt(Hamilton and MacKenzie, 1960). Experiments at 1 and 2 kbars show that the effectof pressure on the solid-solution boundary is negligible (Hamilton, 1961). Thismethod was applied to the Type-I aggregates that show textural evidence of coevalcrystallization of nepheline and carbonate (Fig. 2B). Two analyzed nephelinescontain 73.8 and 82.4 mole% nepheline end-member (NaAlSiO4), 26.2 and17.6 mole% kalsilite (KalSiO4), and 1.9 and 3.5 mole% SiO2, respectively, and

Table 4. Average ice-melting temperatures and salinities of fluid inclusions in carbonateaggregates from the TCR lamprophyres

Aggregatetype

Sample AverageTh (�C)

S.D. AverageTm (�C)

Salinity(NaClequiv. wt%)

S.D.

Type-I St-1=4a 141 2.2 �9.6 13.6 n=aSt-1=4b n=a n=a �7.3 10.9 n=aMy-1=1 204 4.3 �3.3 5.4 1.5Pa-1=1 102 33.1 n=a n=a n=aBkt1=6 130 49.3 �0.2 0.3 n=a

Type-II St-1=1c 111 22.6 �1.0 1.7 2.6St-1=3 128 29.7 �1.0 1.7 4.1

Type-III St-1=6 106 15.1 n=a n=a n=aMy-1=2 158 36.1 �2.43 4.1 2.6Rh-121=3 90 6.9 �0.2 0.3 n=a

Th Homogenization temperatureS.D. Standard deviation of measurementsTm Final melting temperature of ice

490 T. Azbej et al.

Tab

le5

.Sum

mary

of

pet

rogra

phic

,geo

chem

ical

and

mic

roth

erm

om

etri

cch

ara

cter

isti

cso

fdif

fere

nt

carb

onate

aggre

gate

types

inth

eT

CR

lam

pro

ph

yres

Ag

gre

gat

eT

yp

eP

etro

gra

phy

Mic

roth

erm

om

etry

Car

bonat

eC

om

posi

tion

Pro

pose

dO

rigin

Sam

ple

s

Ty

pe-

IG

lobu

lar

shap

e,T

ang

enti

ally

ori

ente

dm

ica

and

amphib

ole

atri

m

Pri

mar

yF

IsT

hb

etw

een

77

and

20

4� C

Dolo

mit

e,ca

lcit

e,m

agnes

ite

Hy

dro

ther

mal

fill

ing

of

ves

icle

so

rre

cryst

alli

zati

on

of

oce

lli

Ad

-2=

III,

Bkt1=6

,M

y-1

,P

a1,

St-

1,

Val

-3

Type-

IIP

oly

gonal

shap

e,O

livin

epse

udom

orp

hs,

no

mic

ao

ram

ph

ibo

leri

m

Pri

mar

yF

IsT

hb

etw

een

95

and

17

2� C

Dolo

mit

eP

seudom

orp

hs

afte

roli

vin

eA

d-2=

X,

St-

1

Ty

pe-

III

Irre

gu

lar

shap

e,n

om

ica

or

amp

hib

ole

rim

,A

l-si

lica

tes

atri

m

Sec

on

dar

yF

IsT

h�

10

4� C

Com

posi

tional

zonat

ion:

calc

ite,

dolo

mit

e,m

agnes

ite,

�an

ker

ite

Car

bonat

exen

oli

ths

and

xen

ocr

yst

sA

d-2=

II,

B€ oo1

-16

,R

h=1

01

2,

Rh

8,

Bk

t1=6

Fls

Flu

idin

clu

sio

ns


suggest a temperature of formation <500 �C. This temperature is assumed to alsoapply to the contemporaneous carbonate in the aggregates.

Genesis of the carbonate aggregates

The genesis of Types-II and -III carbonate aggregates can be readily established onthe basis of the evidence presented here (Table 5). Petrographic and geochemicalcharacteristics of Type-III carbonates support their xenolithic – xenocrystic origin.Their polygonal, irregular shape (Fig. 2E, F) is consistent with their origin asfragments from the conduits of the intruding lamprophyre melt. The other distinc-tive geochemical feature of the Type-III aggregates is that they contain Mg- andFe-rich carbonate phases at the rims, and Ca-rich carbonate cores (Fig. 2E, F).

The geological setting and petrographic and geochemical features of zonedsilicate phases (clinopyroxenes, mica) from lamprophyres indicate an extremelyrapid cooling rate for the dike rocks (Szabo et al., 1993). According to Fisler andCygan (1999), distinct carbonate compositional zoning on a scale of 100’s ofmicrometers, such as observed in Type-III carbonate aggregates, is unlikely tobe reset by chemical diffusion during rapid cooling (1000 �C=m.y.) of melts incontact with carbonates. Thus, we can exclude chemical diffusion for the originof compositional zoning observed in Type-III aggregates (Fig. 2E, F). The compo-sitional zonation in Type-III aggregates (Fig. 2E, F; Table 5) is best explained byreaction between the solid carbonate wall rock and hot lamprophyre melt, resultingin rapid melting and crystallization of the rims of carbonate xenoliths.

The preceding interpretation is consistent with phase equilibria in lamprophyremelts, which have liquidus temperatures �1200 �C at lower crustal conditions(Nemec, 1977; Esperanca and Holloway, 1987). Most lamprophyre melts are almostcompletely crystallized after cooling to 250 to 450 �C below the liquidus tempera-ture (Nemec, 1977; Montel and Weisbrod, 1986). Thus, melts interacting withcarbonate (limestone and=or dolomite) wallrocks in the upper crust must have beenat least 750 �C. Following entrainment of wall-rock fragments, lamprophyric meltsrise quickly towards the surface (Rock, 1991; Szabo et al., 1993) and cool rapidly,leaving little time for complete assimilation of the larger carbonate aggregates.Thus, only the rims of Type-III aggregates show evidence of interaction with thehost melts.

The occurrence of Al-silicates such as those observed at the contact betweensome Type-III carbonate aggregates and the lamprophyre groundmass (Fig. 2F)have been interpreted as the product of partial assimilation of limestone by alkalineand mafic magmas (e.g. Joesten, 1977; Baker and Black, 1980; Joesten et al., 1994;Owens, 2000). The secondary fluid inclusions are likely to represent hydrothermalfluids related to the late-stage crystallization of the lamprophyre, or to fluids infil-trating the dikes after their complete crystallization.

The shape and size of Type-II carbonate is similar to that of unaltered orpartially altered olivine phenocrysts in the lamprophyres (Fig. 2D) and appear tobe metasomatic carbonatization products after olivine phenocrysts. This is consis-tent with observations of Rock (1991) who described widespread late-stage orsubsolidus autometasomatic alteration of primary minerals in lamprophyres dueto their high volatile content. However, primary fluid inclusions in Type-II aggre-

492 T. Azbej et al.

gates are at variance with the proposed gradual carbonatization of olivine. Theolivine must have been completely altered and removed to provide space for pre-cipitation of later carbonate phases by hydrothermal fluids, some of which weretrapped to produce primary fluid inclusions.

Based on the tectonic and stratigraphic settings of the lamprophyre dikeswarms, late stage, secondary process associated with formation of Type-II carbon-ate aggregates occurred in the upper crust (Kubovics and Szabo, 1988). Thus, thecarbonate aggregates are likely to have formed at pressures lower than 5 kbars.Based on the isochores for aqueous (H2O–NaCl) fluid inclusions with salinitiesbetween 5 and 10 wt% and homogenization temperatures below 200 �C (Fig. 5,Table 4), the formation temperature of these fluid inclusions would have beenunder 500 �C if the pressure was lower than 5 kbars (Fig. 5). A notable feature isthat Type-II aggregates occur exclusively in the St-1 and Ad-2 lamprophyre samplewhere the abundance of Type-I aggregates with a similar composition is high(Tables 1, 2), as well. This suggests that the source of fluids responsible for thegenesis of Types-I and -II carbonates might be the same. This is supported by theirsimilar composition (Fig. 3, Tables 1, 2) and by the fact that there is a network ofcarbonate veins connecting Types-I and -II aggregates (Fig. 2G, H).

Petrographic and geochemical features of Type-I aggregates are similar to thoseof aggregates that have been interpreted to be ocelli (e.g. Phillpotts and Hodgson,1968; Ferguson and Currie, 1971). However, most of these authors explained the

Fig. 5. Estimates of P–T formation conditions for carbonate aggregates in the TCRlamprophyres. The shaded field shows the pressure–temperature field for formation ofType-I carbonate aggregates, defined by the isochores (dashed lines) for lowest and highestmeasured homogenization temperatures and lowest and highest calculated salinity. Thevertical solid line shows the estimated maximum temperature of formation and thehorizontal line shows the assumed maximum pressure of formation. Also shown the water-saturated carbonatite solidus from Boettcher et al. (1980)


genesis of globular aggregates in alkali mafic rocks as solidified droplets of car-bonate melt that separated from a silicate melt. Formation of the carbonate phasesin the ocelli of TCR lamprophyres, however, cannot be explained by immiscibilityduring crystallization of the felsic melt. However, the vugs that now contain car-bonate could have been formed during such a process.

Primary fluid inclusions in Types I- and -II aggregates show homogenizationtemperatures lower then 204 �C (Fig. 4). Assuming a formation pressure of<5 kbars for the lamprophyres, the formation temperature of the primary inclu-sions would have been lower than the water-saturated carbonate solidus (Fig. 5).This conclusion is also supported by the estimated temperature range (lower than500 �C) for the formation of silicate phases in the ocelli rims that show texturalevidence for coeval crystallization with the carbonate phases (Fig. 2B). Most likelyType-I ocelli formed by the following process. First, gas bubbles exsolved from thevolatile-rich, late stage melts to produce vesicles in the magma after much of thegroundmass had crystallized (Foley, 1984; Andronikov and Foley, 2001). The com-monly occurring tangential alignment of phlogopites around aggregates (Fig. 2C)was caused by expansion of gas bubbles in a partially crystallized magma (Phillips,1973). External fluids responsible for the precipitation of Type-I carbonate aggre-gates were transported through the fracture system now preserved as carbonate-filled veins. The important role of externally derived fluids in the formation of theaggregates is also supported by the mixed magmatic-meteoric isotopic compositionof the carbonate aggregates reported by Demeny et al. (1994).

Based on the petrographic and geochemical characteristics of Type-I carbo-nates, one cannot exclude the possibility that they represent recrystallized mag-matic carbonates. This also implies that the fluid inclusions do not provideinformation on the formation of the carbonate aggregates, but rather that the inclu-sions represent fluids that were present during recrystallization. Among the mostdiagnostic tests recrystallization has occurred is the presence of geochemical pat-terns that overprint original compositional zoning (Goldstein, 2003). CL and BSEimaging show homogenous cathodoluminescence textures in Type I aggregates.Such textures are usually not associated with recrystallization, but cannot be usedto conclusively rule out that recrystallization has occurred.

Results of integrated petrographic, geochemical and fluid inclusion studies sug-gest that ocelli previously interpreted as products of carbonate–silicate melt immis-cibility might be better interpreted as the products of hydrothermal processes.

Conclusions

Late Cretaceous lamprophyres from the Transdanubian Central Range (Hungary)contain carbonate aggregates that have been classified into three distinct groups(Table 5). Type-I aggregates are of globular shape, contain primary aqueous fluidinclusions, lack major element zonation, and show tangentially aligned mica at thecontact with the host rock. Based on microthermometric analyses and geother-mometric calculations, these aggregates precipitated from aqueous hydrothermalsolutions to fill vesicles in the crystallized (or partially crystallized) melt. This ob-servation is inconsistent with previous models that explain the genesis of suchfeatures by silicate–carbonate melt immiscibility.

494 T. Azbej et al.

Type-II carbonate aggregates also host primary fluid inclusions but show poly-gonal shape and lack oriented sheet silicates at their rims. Type-II aggregatesformed from hydrothermal fluids, similar to those forming Type-I aggregates,except that the carbonate phases precipitated in spaces previously occupied byolivine phenocrysts.

Type-III aggregates have irregular shape, distinct compositional zoning(increasing Mg-, and Fe-content in the carbonate phases towards the rims), containsecondary fluid inclusions, and contain Al-silicates (clay minerals) at the contactwith the lamprophyre host, instead of a shell of oriented mica as in Type-I. Theseaggregates show signs of reaction with the hot lamprophyric melt and are inter-preted to represent xenoliths and xenocrysts from the wall rock of the magmaconduits.

Acknowledgements

This manuscript has benefited greatly from help of Enik}oo Bali, Kalman T€oor€ook and others atthe Lithosphere Fluid Research Laboratory at E€ootv€oos University, Budapest and the FluidsResearch Laboratory at Virginia Tech. A. H. Rankin and F. Wall are thanked for their criticalcomments on an earlier version of this manuscript. Editorial comments of Benedetto De Vivoare also appreciated.

Financial support for this work was provided by TET Hungarian-American Foundation toCsaba Szabo and Robert J Bodnar (17=MO=01) and by Pro Renovanda Culturae HungariaeFoundation to Tristan Azbej and by a grant from the U.S. National Science Foundation toRobert J Bodnar (NSF Grant EAR-0125918).

This is the No. 24 publication of the Lithosphere Fluid Research Lab of the Departmentof Petrology and Geochemistry at E€ootv€oos University, Budapest.

References

Andronikov AV, Foley SF (2001) Trace element and Nd–Sr isotopic composition ofultramafic lamprophyres from the East Antarctic Beaver Lake area. Chem Geol 175:291–305

Azbej T (2002) Petrographic and geochemical study on carbonates in lamprophyres(in Hungarian). M.Sc. Thesis. Lithosphere Fluid Research Lab, E€ootv€oos University,Budapest, p 102

Azbej T, Szabo C, Bodnar RJ, Dobosi G (2003) A new interpretation of the genesis ofcarbonate aggregates from lamprophyres in Hungary: Are ocelli of magmatic or hydro-thermal origin? Acta Mineral Petrogr Abst. Ser, Szeged, Abstract Series 2: 12

Baker CK, Black PM (1980) Assimilation and metamorphism at a basalt-limestone contact,Tokatoka, New Zealand. Mineral Mag 43: 797–807

Balog A, Read JF, Haas J (1999) Climate-controlled early dolomite, late Triassic cyclicplatform carbonates, Hungary. J Sediment Res 69: 267–282

Bence AE, Albee AL (1968) Empirical correction factors for the electron microanalysis ofsilicates and oxides. J Geol 76: 382–403

Bodnar RJ (1992) Revised equation and table for determining the freezing point depressionof H2O–NaCl solutions. Geochim Cosmochim Acta 57: 638–684

Bodnar RJ (2003a) Introduction to fluid inclusions. In: Samson I, Anderson A, Marshall D(eds) Fluid Inclusions: Analysis and Interpretation, Mineral Assoc Canada, Short Course32, Ottawa, pp 1–8


Bodnar RJ (2003b) Reequilibration of fluid inclusions. In: Samson I, Anderson A, Marshall D(eds) Fluid Inclusions: Analysis and Interpretation, Mineral Assoc Canada, Short Course32, Ottawa, pp 213–230

Boettcher AL, Robertson JK, Wyllie PJ (1980) Studies in synthetic carbonatite systems;solidus relationships for CaO–MgO–CO2–H2O to 40 kbar and CaO–MgO–SiO2–CO2–H2O to 10 kbar. J Geophys Res 85: 6937–6943

Cooper AF (1979) Petrology of ocellar lamprophyres from western Otago, New Zealand.J Petrol 20: 139–163

Dem�eeny A, Forizs I, Molnar F (1994) Stable isotope and chemical compositions of carbonateocelli and veins in Mesozoic lamprophyres of Hungary. Eur J Mineral 6: 679–690

Dunkl I (1991) Application of fission track methods in geochronological methods(in Hungarian). Ph.D. Thesis, Technical University of Miskolc, Miskolc, p 178

Embey-Isztin A, Dobosi G, Noske-Fazekas G, �AArva-Sos E (1989) Petrology of a new basaltoccurrence in Hungary. Mineral Petrol 40: 183–196

Esperanca S, Holloway JR (1987) On the origin of some mica-lamprophyres; experimentalevidence from a mafic minette. Contrib Mineral Petrol 95: 207–216

Ferguson J, Currie KL (1971) Evidence of liquid immiscibility in alkaline ultrabasic dikes atCallender Bay, Ontario. J Petrol 12: 561–585

Fisler DK, Cygan RT (1999) Diffusion of Ca and Mg in calcite. Am Mineral 84: 1392–1399Foley SF (1984) Liquid immiscibility and melt segregation in alkaline lamprophyres from

Labrador. Lithos 17: 17–137Goldstein RH (2003) Petrographic analysis of fluid inclusions. In: Samson I, Anderson A,

Marshall D (eds) Fluid Inclusions: Analysis and Interpretation. Mineral Assoc Canada,Short Course 32, Ottawa, pp 9–55

Goldstein RH, Reynolds TJ (1994) Systematics of fluid inclusions in diagenetic minerals.Society for Sedimentary Geology, Short Course 31, Ottawa, p 199

Hamilton DL (1961) Nephelines as crystallization temperature indicators. J Geol 69:321–329

Hamilton DL, MacKenzie WS (1960) Nepheline solid solution in the system NaAlSiO4–KalSiO4–SiO2–H2O. J Petrol 1: 56–72

Hamilton DL, Freenstone IC, Dawson JB, Donaldson CH (1979) Origin of carbonatites byliquid immiscibility. Nature 279: 52–54

Horvath I, �OOdor L (1984) Alkaline ultrabasic rocks and associated silicocarbonatites in theNE part of the Transdanubian Mts. (Hungary). Mineral Slov 16: 115–119

Horvath I, Daridan�ee TM, �OOdor L (1983) Magnesite bearing dolomitic carbonatite dike rocksfrom the Velence Mountains (in Hungarian). Magy �AAll F€ooldt Int �EEvi Jel 1981-r}ool, 41–44

Joesten R (1977) Mineralogical and chemical evolution of contaminated igneous rocks at agabbro-limestone contact, Christmas Mountains, Big Bend region, Texas. Geol Soc AmBull 88: 1515–1529

Joesten R, Hill J, Van-Horn SR (1994) Limestone assimilation and clinopyroxene productionalong the contacts of a 9 meter alkali olivine basalt dike. Geol Soc Am, Abstracts withPrograms, 26, Killala Bay, Ireland, p 476

Kazm�eer M, Szabo C (1989) Late Cretaceous lamprophyre dikes in the hinterland of theAlpine deformation front. EUG-V, Terra abstract, 4, Strasbourg, p 177

Kubovics I, Szabo C (1988) Geochemical, petrologic and mineralogical study on alkali maficand ultramafic dike rocks from Alcsutdoboz-2 drill hole (in Hungarian). Magy �AAll F€ooldtInt �EEvk LXV: 335–356

Kubovics I, Gal-Solymos K, Szabo C (1985) Petrology and geochemistry of ultramaficxenoliths in mafic rocks of Hungary and Burgenland (Austria). Geol Carpathica 36:433–450

496 T. Azbej et al.

Kubovics I, Szabo C, Gal-Solymos K (1989) A new occurrence of lamprophyre in the BudaMountains, Hungary. Acta Geol Hung 32: 149–168

Kubovics I, Szabo C, Harangi Sz, Jozsa S (1990) Petrology and petrochemistry of Mesozoicmagmatic suites in Hungary and the adjacent areas – an overview. Acta GeodaetGeophys Mont Hung 25: 345–371

Montel JM, Weisbrod A (1986) Characteristics and evolution of ‘‘vaugneritic magmas’’; ananalytical and experimental approach, on the example of the Cevennes Medianes (FrenchMassif Central). Bull Mineral 109: 575–587

N�eedli Z, M. Toth T (2003) Petrography and mineral chemistry of rh€oonite in ocelli of alkalibasalt from Villany Mts., SW-Hungary. Acta Mineral-Petrogr 44: 51–56

Nemec D (1977) Differentiation of lamprophyre magma. Krystalinikum 13: 73–87Owens BE (2000) High-temperature contact metamorphism of calc-silicate xenoliths in the

Kiglapait Intrusion, Labrador. Am Mineral 85: 1595–1605Phillips WJ (1973) Interpretation of crystalline spheroidal structures in igneous rocks. Lithos

6: 235–244Philpotts AR (1990) Principles of Igneous and Metamorphic Petrology. Prentice Hall.

Englewood Cliffs, New Jersey, p 498Philpotts AR, Hodgson CJ (1968) Role of liquid immiscibility in alkaline rock genesis.

XXIII. Int Geol Congr 2, pp 175–188Rock NMS (1991) Lamprophyres. Blackie and Son, Glasgow=London=New York, p 285Szabo C (1985) Xenoliths from Cretaceous lamprophyre of Alcs�uutdoboz-2. borehole, Trans-

danubian Central Mountains, Hungary. Acta Mineral-Petrogr 27: 39–50Szabo C, Kubovics I, Molnar Z (1993) Alkaline lamprophyre and related dike rocks in

NE Transdanubia, Hungary: The Alcsutdoboz-2 (AD-2) borehole. Mineral Petrol 47:127–148

Vichi G, Stoppa F, Wall F (2005) The carbonate fraction in carbonatitic Italian lamprophyres.Lithos 85: 154–170

Authors’ addresses: T. Azbej� (e-mail: [email protected]) and Csaba Szabo (correspondingauthor; e-mail: [email protected]), Lithosphere Fluid Research Laboratory, Depart-ment of Petrology and Geochemistry, E€ootv€oos University, H-1117 Budapest, Pazmany P�eeters�eetany 1=c, Hungary; Robert J. Bodnar (e-mail: [email protected]), Fluids Research Laboratory,Department of Geosciences, Virginia Polytechnic Institute and State University, 4044Derring Hall, VA 24061, Blacksburg, USA; Gabor Dobosi (e-mail: [email protected]),Institute for Geochemical Research, Hungarian Academy of Sciences, H-1112 Budapest,Buda€oorsi �uut 45, Hungary; �Fluids Research Laboratory, Department of Geosciences, VirginiaPolytechnic Institute and State University, 4044 Derring Hall, VA 24061, Blacksburg, USA


genesis of carbonate aggregates in lamprophyres from the ...azbej_min_pet.pdf · virginia tech....

Documents